1. Field of the Invention
The present invention relates to the field of detecting a concentration of a blood constituent, for example sugar in blood flowing through a blood vessel.
2. Related Technology
In order to ascertain a blood picture parameter, such as, for example, a concentration of a blood constituent, blood can be taken invasively. The blood picture parameter can then be ascertained using the taken blood by means of standardized test strips, the electric resistance values of which depend on the concentration of the blood constituent, e.g. blood sugar. By way of example, the respective electric resistance value can be detected using a blood sugar measuring instrument, which carries out a DC current resistance measurement for detecting an electric resistance value of a test strip. The resistance value can be converted into a blood sugar concentration on the basis of a relationship, known per se, between a blood sugar concentration and a resistance value. In order to obtain high detection accuracy, each test strip is provided with calibration data, for example with a reference resistance value or with a corresponding code, as a result of which variations of properties of the test strips can be compensated for. However, a disadvantage of invasive methods is the necessity of taking blood and hence of injuring a patient. Moreover, continuous detection of a concentration of a blood constituent, for example to establish the diurnal variation curve thereof, is complicated. Furthermore, it is not possible to detect a time delay between food being taken and, for example, an increase in the blood sugar accurately by means of the invasive method. Also, particularly in the case of a low concentration of the blood sugar in blood, the time for administering insulin to the patient cannot be ascertained accurately.
For noninvasive ascertaining of a blood picture parameter such as, for example, a substance concentration or a substance composition in the blood, use can be made of microwave-spectroscopic methods. Microwave spectroscopy for detecting blood picture parameters is based on coupling a microwave signal into tissue perfused by blood and detecting a frequency-dependent absorption of coupled-in microwave energy.
The article “Non-invasive glucose monitoring in patients with Type 1 diabetes: A multi-sensor system combining sensors for dielectric and optical characterization of skin”, Biosensors and Bioelectronics 24 (2009) 2778-2784 by Andreas Caduff et al. describes a multi-electrode arrangement for microwave-based ascertaining of a blood picture parameter. The multi-electrode arrangement comprises a plurality of electrode pairs with different electrode spacings, by means of which different penetration depths of microwave signals can be realized. The blood picture parameter is detected by means of an impedance measurement, i.e. by means of a one-port measurement, and is therefore susceptible to errors in the case of possible impedance maladjustments. As a result of different penetration depths, it is sometimes not possible to distinguish between capillary and venous blood, which can falsify the measurement results. In general, a measurement of a blood picture parameter using venous blood is more precise than a measurement of the blood picture parameter using capillary blood because, for example, blood sugar changes in capillary blood are delayed compared to venous blood.
The articles “A microwave frequency sensor for non-invasive blood-glucose measurement”, SAS 2008-IEEE Sensors Applications Symposium, Atlanta, Ga., Feb. 12-14, 2008, by Buford Randal Jean et al. and “Calibration methodology for a microwave non-invasive glucose sensor”, Master's Thesis, Baylor University, May 2008 by M. McClung describe a further electrode arrangement for ascertaining a blood sugar concentration. What is exploited here is that the dielectric properties of blood depend on a blood sugar content. By pressing a thumb onto the microwave sensor, a change in the relative permittivity of the thumb is measured by a detuning of a resonator. However, blood is displaced by the contact pressure of the thumb, and this can lead to falsification of the measurement results. Moreover, the measurements cannot be carried out continuously. The evaluation of the measurement data for ascertaining the blood sugar content moreover depends on the respective patient and is therefore not reproducible in other patients. Moreover, this method does not allow control of the penetration depth of the microwave power, and so it is not possible to distinguish between capillary and venous blood. Furthermore, the change in the relative permittivity is carried out on the basis of a one-port measurement, which is susceptible in respect of maladjustments.